[stats-dev] new ogr-ver-incr method

Chris Hodson nerf at slacker.com
Tue Dec 9 15:10:51 EST 2003


On 09-Dec-2003, Jim C. Nasby wrote:
> > 1. Load day_results with today's work (normalized and de-duped)
> I don't think you should be de-duping. I'd like -verify to know what
> users are slamming us with duped work.
I simply ment that instead of separate rowes, duped work gets rolled up and
increments a counter (as it currently does).  Nothing new in this step from
any of the versions.

> > 6. Insert day_summary into ogr_summary (could really combine steps 5 and 6)
> Yeah, you should combine those, otherwise you're needlessly creating a
> temporary table.
I don't know that this matters, as it's either an explicit temp table or an
system one, but it might as well be all together.

The real issue here is "Do youy think it's worth debugging what we have, or
should I go with a simpler way first that gets us half way to our speed
goal, but can be rolled out sooner (probably)."
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.distributed.net/pipermail/stats-dev/attachments/20031209/48c5bd14/attachment.bin


More information about the stats-dev mailing list